13:32:09 RRSAgent has joined #webfonts 13:32:09 logging to http://www.w3.org/2015/10/13-webfonts-irc 13:32:11 RRSAgent, make logs world 13:32:11 Zakim has joined #webfonts 13:32:13 Zakim, this will be 3668 13:32:13 I do not see a conference matching that name scheduled within the next hour, trackbot 13:32:14 Meeting: WebFonts Working Group Teleconference 13:32:14 Date: 13 October 2015 13:32:23 zakim, this meeting spans midnight 13:32:23 I don't understand 'this meeting spans midnight', ChrisL 13:32:31 rrsagent, make logs public 13:32:44 rrsagent, this meeting spans midnight 13:34:48 Present+ Vlad, Chris, Behedad, Garret, Rod, David, Nashan, Raph 13:34:55 Present+ sergeym 13:35:11 Present+ (Zurich Team) 13:35:23 Scribe: ChrisL 13:35:33 Topic: WOFF2 compression improvements 13:37:06 (slides will be online) 13:37:07 Vlad has joined #webfonts 13:37:21 Meeting: WebFonts WG f2f, Brazil 13:39:35 Present+ jfkthame 13:45:45 decompress and reconstructtake about 50% of the CPU time each, so both need to be improved. Improving just one had limitedd effectiveness 13:46:01 recently, much better performance on ARM 13:46:17 16.8% improvement (see slides) 13:46:42 20.8% woff2 speedup on x64 13:48:43 optimising brotli, 2.4x speedup on ARM 13:58:43 doing decompression and reconstruction on separate threads might be faster (see "speculative" on the discussion graph in slides) 13:58:47 Chair: Vlad 13:59:25 Vlad: need to balance download time for extra data vs. decompression speeed gaons. End to end timing. 14:00:05 kuettel: Some people say use woff1 vs woff2 in some cases. But maybe can tune woff2 appropriately for each use case 14:00:35 ... woff1 faster in some cases, which motivated this work 14:00:53 Vlad: click to available font time is important. 14:01:27 kuettel: we have seen incredibly high cache hitrates which removes the transfer time 14:02:17 raph: speed in Mb/s is the quantity that makes sense, even slowest is 30Mb/sec which is 300 mbits/s so thatputs a bound on the network speed whewre transfer time make s adifference 14:02:36 ... on google fibre, send woff1 :) 14:03:25 Vlad: monotype announced a new websitre design, so it is slow until the chache is refreshed with all the new fonts 14:03:55 raph: we want to optimise for device utilisation pov 14:04:34 kuettel - milliseconds in delivery time affect the economic balance 14:06:09 (we need more data on worst case performance) 14:06:59 Vlad: discussed with Kenji, on total time, and these times are 3-6 seconds on some web pages 14:07:12 ... decompression is a tiny fraction of that 14:11:06 ChrisL: brotli ID is onw in the "ISE review" status. This means Independent Specification Editor 14:12:01 https://datatracker.ietf.org/doc/draft-alakuijala-brotli/ 14:13:31 Yergei: Marc said he had completed spec review and built an independent decoder. he is enthusiastic and supportive and has completed review 14:13:41 ChrisL: lets havew that email in the public archive 14:19:40 ChrisL: also a link to the slides and a spreadsheet of the new data on google font corpus, please 14:19:53 rrsagent, make minutes 14:19:53 I have made the request to generate http://www.w3.org/2015/10/13-webfonts-minutes.html ChrisL 14:20:30 topic; agenda 14:20:59 Vlad: spec updates, possible spec modifications based on additional data on htmx and skipping all transforms 14:21:21 rod: not completed that analysis, code is doing unnecessary things so far 14:21:40 Vlad: also a discussion on CTS with Khaled 14:22:00 kuettel: also lunch, from 11:30 on it is available, maybe break at 12 14:22:32 ChrisL: yes, mostly 14:23:09 ChrisL: also agenda item on font/* top level type 14:25:35 don't worry about me ... i'll be hanging around for several hours yet 14:25:53 though i'll come and go a bit i'm sure - just do whatever works for you there 14:26:53 Present+ Zoltan, Jyrki, Evegenii 14:27:23 sergeym has joined #webfonts 15:35:24 15:35:34 topic: spec review, latest updates 15:35:59 Vlad: update, recent discussion on hmtx 15:36:19 rod: khaled said the new tests look fine to him 15:36:36 Vlad: some editorial changes wrt the suggested text from jfkthame 15:36:50 ... accepted with modifications, and some text was not dropped 15:39:49 RSheeter has joined #webfonts 15:39:56 Vlad: first update was para after table entry description, modified the flag description 15:40:01 http://dev.w3.org/webfonts/WOFF2/spec/#table_dir_format 15:42:18 Vlad: if we decide to do untransformed we have three bits available as flags for transform version 15:43:56 (discussion on breaking changes, 000=untransformed would break all existing woff2 fonts) 15:44:35 kuettel: you were looking at making it backwards compat? 15:44:44 RSheeter: that didn't work 15:45:22 does http://dev.w3.org/webfonts/WOFF2/spec/ at least work? 15:45:52 raph: what is the vaslue of making a change that invalidates all fonts. and if not, what do we do for existing. its crufty to define 001 = untransformed for glyf and loca 15:46:52 oh sorry, I thought that didn't work meant the link, lol 15:47:30 raph: needs a really substantial gain, like a huge decrease in filesize, that affects users. a gain in future extensibility does not count here, end users gain nothing 15:47:44 we could handle this by defining a flag in the 'reserved' field of the WOFF2 header, it would mean decoders would need to test an extra flag to know how to interpret the transform version, but would be very cheap to implement 15:48:07 raph: we could do this with magic numbers 15:48:19 ChrisL: that fossilizes the previous design for all time 15:49:02 kuettel: maybe make 000 be the recommended one for typical compression 15:49:36 Vlad: when we decode table directory, every field other than flags is variable length and optional. 15:49:51 ... so it is a bit inconvenient, we rely on specific tag values 15:50:30 ... extending possibility of new transforms, instread of relying on tags but using flag values, it says if it is transformed, and if there is more than one version which one 15:52:28 ChrisL: we could say 000 except "for historical reasons" 001 for glyf and loca 15:52:35 raph: +1 15:52:38 rod: +1 15:53:21 Vlad: for glyf and loca, define 11 as null transform, rest have 00 as nul for all, then easily switch from tags to flags 15:54:04 raph: its a small enough amount of cruft, we can live with it 15:54:50 resolved: do not break backwards compat, go with flags solution and document in spec 15:55:14 action: vlad to update spec for flags, weith glyf and loca treated specially for historical reasons 15:55:14 Created ACTION-188 - Update spec for flags, weith glyf and loca treated specially for historical reasons [on Vladimir Levantovsky - due 2015-10-20]. 15:56:12 Vlad: the existing implementations will need to change from tags to flags to detect presence of transformation 15:56:56 raph: so, tags *and* flags? 15:57:02 Vlad: no 15:57:07 RSheeter: yes 15:57:09 :) 15:57:42 raph and RSheeter are correct :) 15:58:52 (conclusion, only glyf and loca are special; for all other tables the tag is irrelevant) 16:01:06 RSheeter: say in spec not to mark only one of glyf and loca as untransformed, once we introduce the null transform 16:01:20 kuettel: and add a test for that 16:02:15 Vlad: next change was to say for hmtx that null transform is an option 16:02:16 http://dev.w3.org/webfonts/WOFF2/spec/#hmtx_table_format 16:03:00 Vlad: decided to keep the text because it also describes what happens if there is a null transform 16:03:21 ... note added per jfkthame suggestion 16:03:45 Vlad: we need a test for that 16:05:51 kuettel: we have not made the expected gains from disabling transform. 16:05:51 no objection to what vlad's done from me 16:06:11 raph: no harm in the constraint either 16:06:29 Vlad: when transform is applied, these are the constraints 16:07:22 ChrisL: concerned over the "can only be used" in a non-normative note - that is not testable 16:08:50 Vlad: hmtx is required in OT, however CFF has its own numbers and these take precedence. Adobe did not ever update so give hmtx precedence! so we can't resolve that here 16:09:46 ... so in CFF the hmtx table will have different values from those in the CFF 16:09:57 RSheeter: can we say that in the spec? 16:10:28 Vlad: with collections extended to CFF, there may be a cvase with a collection using CFF and different metric for each font, so multiple hmtx 16:10:42 behedad: no because you can't change the LSB 16:10:55 ... (draws on whiteboard) 16:12:09 Vlad: Sairus said they want their implementation updated to follow spec and use hmtx as the preferred one, to also work in collections 16:12:43 s/LSB/lSB/g 16:13:45 raph: language is ambiguous if there are multiple glyf tables *in one file* 16:14:04 ... needs to say "corresponding glyf table" 16:14:12 Vlad: or remove file, just say font 16:15:21 RSheeter: our entire description assumes hmtx belongs to *a* glyf table, in a collection that may not be true 16:15:51 Vlad: hmtx would have to reference all the glyphs in that case 16:16:56 raph: there is a many to many relationshp - imagine a collection with proportional and monospace using the same glyf and two hmtx. only one can be reconstructed 16:18:02 (we agree) 16:24:03 raph: this is now a choice of the encoder, and the spec does not reflect that 16:27:46 sergeym: there exist collections with multiple glyf tables 16:28:39 raph: prefer to not constrain what the encoder does 16:31:23 ... if we remove the word file, so we mean in a font, there is a single table directory. Maybe add language to make explicit that if there are multiple fonts in a file, it must be the case for each hmtx/glyf pair, transform only applied if they match 16:32:10 Vlad: maybe add more explanation, for font collections where hmtx is shared, check that all of them comply else do not transform 16:32:27 RSheeter: check for all the fonts in a collection 16:33:22 Zakim has left #webfonts 16:33:34 ChrisL: prefer for it to be explicit; one is "may transform " and one "must not transform" so it is clear 16:34:07 action vlad to clarify about shared hmtx tables, can only transform is all glyf tables match 16:34:08 Created ACTION-189 - Clarify about shared hmtx tables, can only transform is all glyf tables match [on Vladimir Levantovsky - due 2015-10-20]. 16:34:57 behedad: bounding box can lie about actual xmin, optimisation is toremove it. This one does not say in which way they must match 16:35:16 zakim has left? 16:35:22 Vlad: You refer to glyf transform, not hmtx transform 16:35:34 Zakim has joined #webfonts 16:35:48 zakim, remind us in 5 hours to eat 16:35:48 ok, ChrisL 16:36:11 (more whiteboard, behedad vs. vlad) 16:36:50 vlad uses "it is clear already" It's super effective! 16:38:10 Vlad: if it is dropped we recalculate it first, then go onto the next step 16:38:35 raph: there is a data dependency, cannot calc until you can look at the glyf table 16:38:54 behedad: rather see the hmtx early to know I fill in the lSB 16:39:25 raph: these constrains minimise the decoder memory requirements 16:43:04 raph: so the expected pairng listed in the spec might not be possible in the case of many to one or one to many relationships between glyf and hmtx 16:44:30 raph: do not want additional table order constraints for glyh and hmtx, unlike glyf and loca 16:46:07 raph: need to refer to table directory structure. and the additional constraint of checking all hmtx-glyf pairings 16:46:39 ... normative for decoder reconstruction to use table directory structure to determine that 16:47:33 Topic: Anything other than hmtx, basically 16:49:16 raph: we need all the permutations of one and many glyf and hmtx all in the test suite 16:50:13 action-189? 16:50:13 action-189 -- Vladimir Levantovsky to Clarify about shared hmtx tables, can only transform is all glyf tables match -- due 2015-10-20 -- OPEN 16:50:13 http://www.w3.org/Fonts/WG/track/actions/189 16:52:10 raph: this adds a file format validity test 16:54:36 raph: propose both an authoring test (that it does not transform hmtx) and a ff test (with a n invalidy created transformed hmtx) 16:55:32 action vlad to add conf reqt on AT and FF to test for non-transformable shared hmtx with non-atching metrics in the two glyf tables 16:55:33 Created ACTION-190 - Add conf reqt on at and ff to test for non-transformable shared hmtx with non-atching metrics in the two glyf tables [on Vladimir Levantovsky - due 2015-10-20]. 17:41:16 Topic: CTS planning 18:00:49 https://www.w3.org/Fonts/WG/wiki/Main_Page#WOFF_2.0_Test_Suite 18:16:25 https://www.w3.org/Fonts/WG/wiki/TestPlan20-UserAgent#mustLoadFontCollection 18:17:05 ChrisL: sounds more like a decoder test 18:18:30 browsers have no collection support currently 18:19:47 raph:lower levels of chrome font stack do support collections. ski supports collections 18:19:58 ... ask Kenji 18:20:08 ... chromium bug exists 18:25:16 was moved to decoder tests 18:25:28 https://www.w3.org/Fonts/WG/wiki/TestPlan20-Decoder#mustLoadFontCollection 18:29:55 https://www.w3.org/Fonts/WG/wiki/TestPlan20-UserAgent#mustCheckLSBFlags 19:33:38 https://www.w3.org/Fonts/WG/wiki/TestPlan20-AuthoringTool#mustEliminateLSBs 19:39:58 Topic: font top-level type 19:42:46 Chris has to make a new ID before the IETF Yokohama meeting, to be considered by the APP Area directors at that meeting for adoption by the App Area WG at IETF. 19:42:54 Deadline is the 19th 19:43:12 Then we replace the current appendix with one that just has the font/woff2 type defined 19:44:31 Vlad has joined #webfonts 19:44:35 this relates to action-172 19:45:21 adjourned 19:45:27 rrsagent, make minutes 19:45:27 I have made the request to generate http://www.w3.org/2015/10/13-webfonts-minutes.html ChrisL 21:35:49 ChrisL, you asked to be reminded at this time to eat 23:34:55 jdaggett has joined #webfonts 00:43:41 Zakim has left #webfonts 01:07:48 ChrisL has joined #webfonts 02:01:22 ChrisLilley has joined #webfonts 02:02:26 Chris_Lilley has joined #webfonts